This proposal is aimed at studying mechanisms that drive compulsion-like alcohol drinking (CLAD). Compulsive alcohol intake is characterized by drinking that persists even when alcohol is paired with adverse social, legal and physical consequences, and this consequence-resistant intake is a major obstacle to treating alcohol use disorder (AUD). Thus, we have pioneered the use of rat models to identify brain circuits that underlie CLAD, where drinking continues even when alcohol is paired with aversive stimuli. However, little is known about brain circuits that promote CLAD, especially brain areas that process aversion and allow alcohol drinking to continue in the face of adverse consequences. We previously discovered that the nucleus accumbens core (Core) is a critical region that drives CLAD. The Core has two main types of neurons, which have differential dopamine receptor expression that helps define them as D1-cells versus D2-cells. Although Core D1-cells promote while D2-cells inhibit many addiction-related behaviors, including compulsive cocaine intake, nothing is known about Core D1- and D2-cell activity during alcohol drinking, with or without negative consequences, or the importance of Core D1- or D2-cells for driving CLAD. Since D1-cells promote many behaviors, we predict that increased D1-cell activity will support CLAD. In contrast, D2-cells often promote avoidance of aversive states. Thus, we propose that Core D2-cell activity will be reduced during CLAD, which may be crucial for allowing intake in the face of aversion to persist. Also, we find that CLAD has particular focus on initiation of alcohol-drinking bouts, and we propose that D1-cell activity will be important at the onset of CLAD bouts. To test these possibilities, we use novel genetically-modified rat lines to specifically target Core D1- and D2-cells. Aim 1 utilizes state-of-the-art fiber photometry and the calcium activity indicator GCaMP6f to examine cell-specific activity changes during alcohol drinking. We predict increased D1-cell activity and decreased D2-cell activity during CLAD, relative to alcohol-only intake. Importantly, if Core cells promote CLAD by processing aversion, then activity changes should be larger when aversion during CLAD is greater. Aim 2 uses precisely-timed optogenetic inhibition of Core D1- or D2-cells to determine whether D1-cells promote and D2-cells inhibit CLAD, relative to alcohol-only drinking, and whether these cell types play a particular role at the onset of CLAD. We also collaborate with Dr. Saleem Nicola to attempt to use video tracking to predict when licking will occur. This would allow optogenetic inhibition before the onset of licking, so that we can uncover the importance of D1- and D2-cells when preparing to drink despite adversity. Our studies will provide important and novel information about how specific activity changes in Core cells may process aversion, and thus are critical for the expression of CLAD.